Understanding and Managing Leakage in Forest-Based
Greenhouse Gas Mitigation Projects
By Reimund Schwarze, John O. Niles, and Jacob Olander
Prepared for the Nature Conservancy
With Support from USAID
i
ii
Copyright  2002 The Nature Conservancy. All rights reserved.
This paper has been commissioned of independent specialists by the Nature Conservancy to
contribute information and perspectives on key issues relating to forests and climate change.
The opinions expressed herein are those of the authors and do not necessarily reflect the
views of the Nature Conservancy.
This publication was made possible through the support provided by the Bureau for Economic
Growth, Agriculture and Trade, U.S. Agency for International Development, under the terms
of Award No. LAG-A-00-00-00019. The opinions expressed herein are those of the author(s)
and do not necessarily reflect the views of the U.S. Agency for International Development.
A previous version of this paper appeared in Philosophical Transactions: Mathematical,
Physical and Engineering Sciences, Volume 360, Number 1797, Pages 1685-1703, The Royal
Society, London (DOI: 10.1098/rsta2002.1040).
Reimund Schwarze is at the Institute for Environmental Economics at the Technische
Universität Berlin. He can be contacted at [email protected]
John-O Niles is in the Energy and Resources Group at the University of California, Berkeley.
He can be contacted at [email protected]
Jacob Olander is Director of EcoDecisión, based in Quito, Ecuador, He can be contacted at
[email protected]
iii
PREFACE
Forests have a unique, three-fold relationship to global climate change: They are
simultaneously at risk from the effects of climate change, while being part of the cause and
part of the solution. Diverse climate models indicate that many forest ecosystems will face
future changes in temperature and rainfall regimes, increases in the extent and severity of
forest fires and other factors that may result in broad shifts in forest distribution and
composition. At the same time, forests are a source of greenhouse gases. Some 20-25% of
global CO2 emissions are the result of deforestation and land-use change, primarily in the
tropics – home to a majority of the world’s biodiversity. Finally, conserving and restoring
forests can make a significant contribution to reducing or mitigating greenhouse gas
emissions. Well-designed and implemented projects that reduce the rate of deforestation or
increase the rate of CO2 uptake in new vegetation can yield real, measurable, long-term
climate benefits. While no substitute for needed reductions in fossil-fuel consumption, these
projects can also provide additional benefits to local community development and biodiversity conservation.
The compromises reached in the latest rounds of climate-change negotiations under the
United Nations Framework Convention on Climate Change (UNFCCC) and the Kyoto
Protocol recognize these important roles that forests can play. In the first commitment period
under the Kyoto Protocol (2008-2012) some land use, land-use change and forestry
(LULUCF) activities may be counted towards part of industrialized countries’ obligations to
reduce their net greenhouse gas emissions, both within their borders and internationally
through Joint Implementation and the Clean Development Mechanism. Although the current
United States administration has distanced itself from the Kyoto Protocol, LULUCF projects
have been explored since at least the early 1990s by private companies, federal agencies and
non-governmental organizations as a useful tool for addressing climate change. Though the
shape of future US responses to climate change is as yet to be defined, it is likely that
LULUCF activities will play an important role.
LULUCF activities have been accepted as legitimate elements in the tool kit that policy
makers and project developers have at hand to deal with climate change. If these are to
generate the sorts of real results that are an environmental necessity in the face of climate
change, then they must be underpinned by solid rules, rigorous accounting and transparent
monitoring. This is particularly important because if LULUCF projects result in mere
additional emissions rather than real reductions, as some fear, the result will be a relative
increase in the severity of global climate change, and increase pressure on the world’s forest
resources.
The Nature Conservancy, in both its international and domestic programs, has been engaged
for over a decade in exploring options for using forest conservation and forest restoration to
simultaneously achieve climate change mitigation and biodiversity conservation. TNC,
working with local partners in Belize, Bolivia, Paraguay, Guatemala, Dominican Republic
Brazil and Peru has developed and/or implemented a series of demonstration projects. These
projects have served to generate a wealth of practical experience, highlighting the special
challenges that these sorts of projects imply as well as demonstrating, in practice, that with
rigorous monitoring and careful design, effective responses can be put in place.
TNC has been particularly active in Latin America, a region with a priceless natural heritage
of biodiversity as well as alarming rates of forest loss. In seeking to reconcile the often
apparently contradictory pressures of human needs and biodiversity conservation, of
iv
economic development and environmental quality, many governments, organizations and
communities have seen international investment in climate change mitigation projects as one
possible solution. Policy makers, analysts, NGOs and project developers throughout Latin
America are working through the devilish details of making LULUCF projects work and
work well.
Much of the fine print that still needs to be spelled out involves often inter-related issues such
as permanence, leakage, scale, baselines, additionality, and sustainable development criteria.
In 2001, TNC commissioned this series of papers by leading experts in their fields as part of
an effort to build capacity on climate change in Latin America sponsored by the United States
Agency for International Development (SUAID). The purpose was to shed light and provide
additional information about these issues to policy makers as they strive to hammer out
agreements that will allow for workable projects with measurable, positive consequences for
the global atmosphere. During workshops with Latin American and other experts in climate
policy and science, three issues were identified as particularly worthy of further exploration
since they are key factors in determining the environmental integrity as well as the viability of
projects: permanence, leakage and scale.
In this paper, Reimund Schwarze, John O. Niles and Jacob Olander provide an overview of
leakage, the risk that emissions may be displaced in time or space outside LULUCF project
boundaries, resulting in diminished greenhouse gas benefits. They also summarize some of
the approaches used by project developers and proposed by analysts for effectively managing
or accounting for leakage, to ensure that projects produce real, measurable greenhouse gas
benefits.
Two other papers, in this series, address the issues of permanence and scale.
In their paper on scale, Christiaan Vrolijk and John O. Niles explore the possible trajectories
of supply, demand and price under different scenarios for the global emissions market—and
in particular what impact inclusion of different LULUCF activities might have. In a market
still plagued by uncertainties as policies and rules evolve, they seek to describe how LULUCF
activities might affect the supply of GHG offsets, the demand for projects from other sectors
and the consequences for the market price of offsets. (Available at
www.nature.org/aboutus/projects/climate/docs ).
In a paper on permanence, Pedro Moura Costa describes some of the innovative accounting
frameworks that have been proposed to address the possibility that carbon in forests may not
be permanently stored. Decisions by project developers or national authorities, or
circumstances beyond the project developer’s control (both natural events such as fire and
hurricane, as well as human activity such as illegal logging) may result in the future release to
the atmosphere of carbon held in forest biomass. A variety of methods might be adopted by
policy makers to account for real climate benefits, even where projects are non-permanent.
(Available at www.nature.org/aboutus/projects/climate/docs ).
It should be noted that TNC commissioned these papers by independent experts to provide
cutting-edge perspectives on these critical issues. The results should not be construed as
institutional positions of TNC, but will ideally contribute to the lively and important policy
debate around these issues in Latin America and globally.
Jacob Olander
v
Understanding and Managing Leakage in Forest-Based
Greenhouse Gas Mitigation Projects
By Reimund Schwarze, John O. Niles, and Jacob Olander
Prepared for the Nature Conservancy
With support from USAID
vi
Understanding and Managing Leakage in Forest-Based Greenhouse Gas
Mitigation Projects
By Reimund Schwarze, John O. Niles, and Jacob Olander
Executive Summary
Activities that increase forest cover or decrease deforestation can help reduce atmospheric
carbon dioxide levels. Concerns have been expressed, however, that land use, land-use
change and forestry (LULUCF) projects may only produce greenhouse gas benefits that are
illusory due to a phenomenon known as “leakage”. Leakage is the unanticipated decrease or
increase in greenhouse gas benefits outside of a project’s accounting boundary resulting from
the project’s activities.
Leakage can potentially be significant compared to the scale of planned GHG changes in
mitigation projects. Thus, leakage constitutes a key challenge to sound climate change policy
formulation. In this paper we review the literature on leakage and pay special attention to
LULUCF projects in developing countries.
Leakage is complex and poorly understood
Leakage is an intricate and diverse phenomenon. Market impacts, people moving from place
to place, ecological feedbacks and product life cycle changes are some of the ways leakage
can be manifested.
To illustrate: an avoided deforestation project may cause activity-shifting leakage (people
leaving a project area to go cut trees elsewhere) or market leakage (less timber available due
to the project, more pressure to cut elsewhere). These two types of leakage are the most
commonly cited, and are often perceived as negative (resulting in more emissions or less
sequestration – that is, more atmospheric greenhouse gases). Both of these processes are
intricate, difficult to monitor and complicated by many outside influences.
At the same time this hypothetical project could have unintentional consequences that lead to
more greenhouse gas abatement (positive leakage). One protected forest may help adjacent
forests stay healthy (an example of ecological leakage) or industries may re-gear production
methods to be less polluting as a result of the forest project (life-cycle leakage). For different
projects the relative magnitude of these types of leakage, both positive and negative, will
vary.
Leakage is not restricted to LULUCF projects
Leakage can arise from specific projects as well as policies, it can be positive or negative and
it can occur in any type of mitigation activity. While some modeling evidence shows that
LULUCF activities are at a greater risk for leakage than other sectors of the economy, results
are speculative at this point. For instance, fossil fuel leakage has been estimated in the range
of 4-40% of the original project carbon offsets, a range similar for LULUCF leakage (~0100%). However, no study has shown that real-world leakage is more pronounced in certain
projects types than others. Most modeling studies of LULUCF leakage have focused on
vii
market leakage and none are based on realistic implementation rates of projects in developing
countries.
Leakage in LULUCF projects
Two broad categories of LULUCF activities are considered: conservation projects and
reforestation/aforestation projects.
Conservation projects
In projects that avoid deforestation or modify forest management practices, activity-shifting
leakage risk will depend on project design and local conditions. If a conservation project’s
design does not address underlying drivers of deforestation, activities may shift outside
project boundaries. This may be particularly true in areas where accessible forests are nearby
and where the activity curtailed by a project is mobile. Conversely, well-designed projects
and certain circumstances (e.g., all other nearby forests have been removed or the area is
highly inaccessible) seem to be intrinsically less prone to activity-shifting leakage. Protecting
native forests will often likely generate positive ecological leakage.
Reforestation/Aforestation projects
The magnitude of activity shifting from reforestation or aforestation projects will depend
(again) primarily on local conditions and the project design. If carried out on lands with other
productive uses, tree-planting projects could be especially prone to activity-shifting leakage.
Meanwhile, projects that plant trees on degraded lands with little or no other productive use
will likely produce less leakage.
Market leakage from reforestation/aforestation projects might conceivably be either positive
or negative. For example, large-scale timber plantations could depress timber prices. This in
turn could reduce the incentive to establish new timber plantations elsewhere. Equally
plausible, timber plantations might reduce timber prices and lessen harvesting on other natural
forests. Whether market forces will lead to a decline in forest harvesting (positive leakage) or
the abandonment of plantations (negative leakage) will depend on local market conditions and
circumstances.
In general, small-scale reforestation/afforestation projects focused on environmental
restoration and/or local community needs should have more positive leakage profiles than
large-scale commercial plantations, mainly because they aim to minimize the drivers and
scale of potential leakage.
Options for Responding to Leakage
At the project level, tools exist for preventing leakage, detecting leakage, and adjusting for
leakage. Options for responding to leakage at the project level include: site selection, project
design, leakage contracts, monitoring. Good project design and careful site selections will
minimize negative leakage and maximize positive leakage. Projects that integrate activities
(forest conservation, forest restoration, community development, plantations, etc.) will
probably be more successful in reducing leakage (and overall). Leakage contracts can specify
actions to deter activity shifting and/or to encourage desired actions. Monitoring can be used
to estimate leakage, and once leakage is detected, project carbon offsets should be adjusted
accordingly.
viii
Other broader policy tools can be used to address leakage. These include: discounting,
project-eligibility criteria, and the use of “aggregate baselines”. Baselines, the development
of national, regional or sector ‘standards’ have been one proposed way to address leakage.
This tool is hampered by poor background information for many developing countries as well
as political opposition. Project-eligibility criteria and discounting (or adjustment coefficients)
have also been proposed for grouping project types together and making broad
generalizations and adjustments.
At the broadest level, policy makers also have several other options including a cap on
certain types of mitigation and a balanced portfolio of mitigation types. Negotiators to the
Kyoto Protocol have agreed to limit the volume of carbon offsets that can be achieved
through LULUCF and/or in developing countries. Although these restrictions were not set
primarily to address the risk of leakage, limiting the scope and number of projects may
effectively keep in check the impact of projects on local and/or global timber markets
respectively, and thus limit market leakage. Nevertheless, given the level of the cap, it is not
clear that it will actually constrain projects, nor market leakage.
Combining a “correct” number of avoided deforestation projects with plantations in a
balanced portfolio could result in counter-balancing amounts of timber supply restrictions and
expansions. However, in practice a balanced portfolio approach would be difficult to
accurately implement and would also need to compensate for differences in carbon flux
timings between project types and other factors.
Leakage is a significant, but by no means insurmountable, risk for climate change mitigation.
Both project-level and macro-level approaches can effectively manage leakage. Many pilot
LULUCF projects are already using many of these tools on the ground, although, not for long
enough periods to make robust conclusions.
There is a way to combine project-based calculations and general principles in a way that
rewards positive actions in terms of leakage. A decision-tree framework (as proposed by
Aukland et al, 2002) for appraising leakage would be able to combine many of the positive
aspects of most of the possible tools for addressing leakage. Policy makers can abbreviate
leakage management based on first-order principles, project appraisal, monitoring,
discounting and other techniques. A decision-tree framework could allow certain types of
projects easier certification while placing higher standards on projects that raise ‘red flags”.
Doing so would appropriately combine general principles that can be inferred about leakage
risks with project-specific evaluations.
Understanding and Managing Leakage in Forest-Based Greenhouse Gas
Mitigation Projects
Reimund Schwarze, John O. Niles, and Jacob Olander
1
INTRODUCTION
ix
Projects that increase forest cover or decrease deforestation can help reduce atmospheric
carbon dioxide. This principle underpins the inclusion of certain land use, land-use change
and forestry (LULUCF) activities in both the United Nations Framework Convention on
Climate Change (UNFCCC) and the Kyoto Protocol. Concerns have been expressed,
however, that LULUCF projects may only produce greenhouse gas benefits that are illusory
due to a phenomenon known as “leakage”.
The most frequently cited example of leakage is a LULUCF forest protection project to
reduce greenhouse gas (GHG) emissions. The leakage concern is that forest protection could
cause deforestation to move from the protected forest to a nearby forest, with no net GHG
benefit. Although this example pertains to the LULUCF sector, project activities in virtually
all sectors of GHG mitigation (energy, transportation, etc) have the potential to cause leakage.
As such, leakage constitutes a key challenge for policy makers and project developers to
address in forthcoming climate change policy formulation.
Leakage would not be of great concern if every country measured every GHG flux in its
borders. Any unintended consequences (leakage) of a climate change project or policy in one
area would be registered in another area’s GHG accounting system. But under the principle of
“common but differentiated responsibilities” (Art. 3 of the UNFCCC), developed nations
under the Kyoto Protocol assume binding limits on their GHG emissions, while developing
countries do not. This differentiated approach inevitably creates the risk of leakage since the
world is increasingly economically interconnected; changes in economic activity and prices
reverberate locally and around the world. Without comprehensive monitoring, emissions
reductions by a project in one area may be measured, while project-induced emissions
increases outside project boundaries may not be measured.
This paper analyses leakage in the LULUCF sector since, rightly or wrongly, leakage risk is
perceived as more severe for this sector than for energy, transport or industry. Our analysis
emphasizes, but is not limited to, the use of LULUCF projects in the Kyoto Protocol’s Clean
Development Mechanism (CDM). Effectively managing leakage will also be necessary for
projects and policies that evolve independent of, or in parallel to, the Kyoto Protocol. 1
2
WHAT IS LEAKAGE?
The Intergovernmental Panel on Climate Change (IPCC) Special Report on LULUCF defines
land use leakage as "…the indirect impact that a targeted land use, land-use change and
forestry activity in a certain place at a certain time has on carbon storage at another place or
time” (IPCC 2000, section 2.3.5.2, p. 71). In another section of this report the IPCC defines
leakage as the “unanticipated decrease or increase in GHG benefits outside of the project’s
accounting boundary … as a result of the project activities.” (ibid., section 5.3.3, p. 246).
Various types of leakage are listed in Table 1.
Table 1: Types of leakage
Causes
Effects
1
Project
Policy
Positive
As may be the case if the United States remains opposed to the treaty and forges other greenhouse gas
partnerships.
Mechanisms
Scales
Sectors
2.1
Negative
Activity shifting
Market effects
Life cycle effects
Ecological
Local
Regional
National
Global
Energy (fossil fuels)
LULUCF (biomass)
Causes: projects or policies
Although leakage is often thought of primarily as a project-based concern, unintended GHG
fluxes can occur with the adoption of regulations or policies. The Kyoto Protocol as currently
interpreted could push certain commercial forestry operations to relocate to developing
countries where they would be unencumbered by GHG liabilities (Niesten et al. 2001). The
same concern also affects energy-intensive industries (Wiener 1997).
2.2
Effects: positive or negative
Leakage is often considered undesirable or bad; in the case of GHG fluxes, “bad” meaning
more emissions or less sequestration. In this paper, this type of leakage is called negative
leakage and would ideally be avoided. There are also situations where unintentional outcomes
may be positive (more emission reductions, more sequestration). For example, emissionsreducing activities may be adopted voluntarily outside project boundaries, as has apparently
been the case with reduced-impact logging techniques in a carbon-offset project sponsored by
the New England Power Company in Sabah, Malaysia (UtiliTree 2000).
2.3 Mechanisms
Unintended GHG fluxes arise through two principal avenues: (IPCC 2000; Brown et al. 1997;
Vine et al. 1999, SGS no date given).
Activity shifting: A project or policy can displace an activity or change the likelihood of an
activity outside the project’s boundaries. One example of negative activity-shifting leakage
would be a plantation project that displaces farmers and leads them to clear adjacent forests.
Market effects: A project or policy can alter supply, demand and the equilibrium price of
goods or services, causing an increase/decrease in emitting activities elsewhere. For example,
if a large forest conservation project reduces the local timber supply so that demand is unmet,
this may increase prices and pressures on forests elsewhere.
In economic terms, market-driven leakage is mediated by a change in the price of goods;
whereas activity shifting is when human or other capital changes location. Importantly, these
two leakage types may in some cases be inversely related (notably in the case of forest
conservation). If a project displaces people and activities to adjacent areas, market leakage
may diminish. This is because activity-shifting leakage moves economic activity while market
leakage occurs through net changes in production for a given regional distribution of
activities.
Two other types of leakage bear mentioning.
•
•
2.4
Life-cycle emissions shifting: Mitigation
activities increase emissions in upstream or
downstream activities (e.g. a forest
conservation project leads to increased road
traffic from tourists or a reforestation
project increases the operation of machinery
creating fossil-fuel emissions);
Ecological leakage. Ecological leakage is a
change in GHG fluxes mediated by
ecosystem-level changes in surrounding
areas. In an example of positive ecological
leakage, stopping deforestation can prevent
carbon emissions in forests adjacent to the
intended protected forests. An example of
negative ecological leakage would occur if
a carbon plantation introduces a pathogen to
surrounding forests, leading to their decline
and a net carbon release to the atmosphere.
The magnitude of ecological leakage
compared to other types of leakage has not
been studied.
Scale: local to global
POSITIVE ECOLOGICAL LEAKAGE AND
TROPICAL FOREST CONSERVATION
Within project areas, tropical forest protection results in
reduced emissions and at times, sustained uptake of
carbon. Tropical forest conservation can also lead to
increased sequestration or less emissions (positive
ecological leakage) in areas outside the project
boundary.
Deforestation in one area can lead to “edge effects” on
forests left standing. Edge effects are changes in a
forest community occurring at the border of forests left
standing. Edge effects are often damaging to the
remaining forests. One study showed that deforestation
caused tree mortality in the area deforested and
contributed to substantial biomass declines, and
therefore greenhouse gas emissions, in surrounding
forests (Laurance et al, 1997) Protected forests in a
project area may also maintain favourable microclimatic conditions that aid the resiliency of other
regional forests (Lawton et al, 2001).
Core protected forests also can support pollinators and
insectivores (such as bats) that make nearby agricultural
areas more productive. Maintaining forests can also
control flood and erosion. These factors may negate the
need for new land conversion. Thus, there are numerous
ecosystem feedbacks from forest conservation that may
secure secondary benefits that ultimately can reduce
emissions or increase uptake in areas outside the strict
project boundary.
Leakage can manifest itself at various scales. If
a project or policy alters local prices or behaviors, then local markets and activities may be
changed. Though it is unlikely that any single project would significantly alter world timber
or crop prices, the sum total of many similar individual projects might conceivably impact
global markets. Similarly, ecological impacts of avoiding deforestation from any single
project will not likely alter global circulation patterns. However, numerous projects to stop
deforestation could theoretically serve to maintain historical global circulation patterns
(McGuffie et al. 1995).
2.5
Sectors: fossil fuel or biomass
Leakage can occur in all sectors of the economy where GHG mitigation is conducted. For
convenience sake, one can divide the economy into two sectors – parts of the economy that
produce primarily fossil fuel emissions (energy, transport, industry) and parts that produce
emissions from some form of biomass (vegetation, forests, soils, etc). LULUCF projects are
often perceived as being particularly leakage-prone. Energy projects, however, also face risks
of activity-shifting and market leakage that will need to be considered at the policy and
project level.
Globally, it has been argued that the Kyoto Protocol may drive some energy-intensive
industries to relocate to developing countries. Unencumbered by the constraints and costs of
GHG emissions limits, this form of international leakage could undermine the apparent
reductions. At the project level, clean-energy projects may displace capital stock or change
relative prices of fuels, and cause leakage of emissions. A climate change mitigation project
that builds relatively clean-burning natural gas plants (an existing market trend in many
countries) could increase the price of natural gas in this area. This in turn could drive other
investments to use coal or oil for new generating capacity, negating some, but not necessarily
all, of the GHG benefits of the gas plant.
Estimates of market leakage for forestry activities (see Table 3) are generally higher than
market leakage estimates in the energy sector of 4-15% (Chomitz 2000, p.9). Other studies,
however, set energy sector market leakage at a comparable level. For example, Sedjo 2000
mentions estimates of up to 40% for the energy sector.
3
LEAKAGE AND FORESTRY PROJECTS
Several types of forestry projects can mitigate GHG accumulation in the atmosphere, and
each may be subject to different kinds and degrees of leakage. For simplicity’s sake we divide
LULUCF projects into two broad categories in our analyses2:
•
•
Projects that avoid carbon dioxide emissions by avoiding deforestation or reducing forest
degradation (“conservation” projects). Included in this broad category would be a wide
range of activities, from outright conservation to reduced impact logging and improved
forest and fire management.
Projects that increase carbon sequestration, removing atmospheric CO2 by growing trees
on previously forested areas (reforestation), or areas that have not historically supported
trees (afforestation).
For each of these broad project types we will examine the kinds of leakage that might occur.
Our results from a review of the literature on the types of leakage for these broad project types
are summarized in Table 2. Where possible, we also discuss the magnitude of potential
leakage. Our review of the scale of concern is hampered by the fact that most LULUCF
projects have not been operational long enough to allow detailed examination. The Noel
Kempff Mercado Climate Action Project in Bolivia is the project with the most
comprehensive leakage analysis thus far (Report pending, Winrock International 2002). What
modeling has been done has been primarily on market effects of global scale carbon
sequestration (see Table 3).
2
We do not discuss projects that use biomass to substitute for fossil-fuel intensive products or
processes (e.g. biofuels) as a separate category. Emissions reductions from displaced fossil-fuel
emissions pertain essentially to the energy sector, while biomass plantations as a source of supply can
be considered under reforestation and afforestation.
Table 2: Leakage and forest projects in developing countries
Project Type
Type of activity
Drivers of
leakage3
Leakage
mechanisms
Leakage
Scale
Conservation1
Replace or prevent
subsistence
agriculture
Replace or prevent
commercial
agriculture
Stop commercial
logging
Subsistence
needs (+ or -)
Stop forest
products harvest
for local needs
Reduced impact
logging, enhanced
forest mgmt.
Ecosystem
restoration
Subsistence
needs (-)
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Activity shift
Ecological
Market
Local
Local
Local
Local/Global
Local
Local/Global
Local/Global
Local
Local
Local
Local
Local
Local
Local
Local/Global
Local
Local
Local
Local
Local
Local
Local
Local
Local/Global
Reforestation
&
afforestation2
Small-scale
reforestation for
local needs
Commercial
plantations
Cash crop
markets (-)
Timber markets
(-)
Technology
transfer (+)
Resource and
land availability
(+ or -)
Land availability (-) timber
supply (+)
Timber markets
(+ or -), land
availability (-)
Positive (+)
negative (-)
neutral (N);
-/N/+
+
-/N/+
-/N/+
+
-/N/+
+
+
+
+
N
+/N/+
+/N/N/+
N
+/N/+/N/+/N/-
1) Excluded from the Clean Development Mechanism for first commitment period (2008-2012) of the Kyoto
Protocol in negotiations as of November 2001.
2) Allowed in the Clean Development Mechanism with limitations.
3) Considering only socio-economic drivers of leakage, not ecological leakage drivers. A (+) sign refers to a
driving factor for positive leakage and a (-) sign refers to a driver of negative leakage.
3.1
Conservation Projects
Land use and land-use change projects may reduce emissions from forests by avoiding
deforestation or through the introduction of management practices that lead to increased
carbon storage in forests (for example, by increasing rotation length and reducing harvest
intensity). Avoided deforestation and proactive forest management have been specifically
excluded as eligible project types under the Kyoto Protocol’s Clean Development Mechanism
(CDM) for the first commitment period (UNFCCC 2001). Comprehensive efforts to curb
global warming will eventually need to promote these tools since deforestation and forest
degradation, mostly in the developing world, constitute some 20-25% of global GHG
emissions (Schimel et al. 1996).
Deforestation results from a mix of economic, social and political causes that vary from site to
site. In general, the primary proximate causes of deforestation in the tropics are logging and
conversion to agriculture or grazing. These activities span a spectrum from subsistence
agricultural, to production of globally traded commodities (e.g., vegetable oil, wood pulp,
cacao, rice etc). Behind the proximate deforestation causes are driving forces such as policies,
attitudes and institutions that influence production and consumption (Turner et al. 1993).
Deforestation is often a complex process influenced by cultures, markets, government
policies, property rights and local politics. Deforestation is often driven by rural poverty and
basic needs such as food, shelter, and fuel. Effective leakage control measures will need to
address both the proximate causes (land-use changes) as well as underlying forces (poverty,
land tenure, etc) of deforestation on a case by case basis.
The magnitude of activity shifting leakage will vary greatly across conservation projects. If
neighboring forested lands are easily accessible and the displaced activity is mobile, activityshifting leakage is likely. Where forested land is not readily available (in regions where land
use is already consolidated) or where capital or labor are not mobile, the risk of activityshifting leakage may be quite low. Projects that improve forest management rather than
eliminate forest harvest altogether will be far less vulnerable to activity-shifting leakage.
When new GHG-friendly ideas or technologies are economically competitive, projects may
cause positive leakage. For instance, if reduced impact logging (RIL) is more profitable for
timber companies than traditional techniques, positive activity-shifting leakage could easily
occur as RIL disseminates beyond the project boundary.3
The risks of market leakage will depend on the nature of the market and the scale of any
project. While a single conservation project is unlikely to affect global markets, it may affect
local markets. In most cases, leakage risk will be less than 100%. For every forest protected,
it is unlikely that there will be an equal expansion in production into other forest areas.
Factors such as intensification, product substitution and reuse will likely replace some of the
displaced output.
Little analytical work has been conducted on the magnitude of possible leakage from
individual forestry projects. Most modeling has evaluated market effects of global scale
carbon sequestration, and only two studies specifically address forest conservation (see Table
3).
In terms of leakage, conservation projects that eliminate logging activities could constrain
supply and lead to increased demand for wood from new sources. Sohngen et al. (1999)
studied the effect of an increased worldwide timber demand on harvesting rates in
inaccessible northern forests such as Siberia, and in tropical South American and African rain
forests. Their study suggests that higher timber prices would produce a stronger response in
harvesting northern inaccessible forests than it does in additional logging in the tropics
despite similar characteristics (low productivity, long rotation periods). Moreover, they show
that an increasing timber price will more likely lead to an intensification of forest
management in existing temperate zone forests (e.g. shorter-rotation periods) or to the new
establishment of plantations in emerging subtropical regions (i.e., South and Central America,
Africa and the Asia Pacific region). The latter results from greater harvest productivity in
temperate zones and subtropical plantations compared to logging of inaccessible tropical
forests. These results illustrate the enormous complexity of estimating leakage. Projects that
stop deforestation caused by logging can reduce timber supply, raise prices and lead to more
logging in temperate forests (negative leakage). Or, they can cause more plantations to be
established (possibly positive leakage). Interestingly, the former scenario is not a leakage
problem since it displaces emissions to where they would be accounted, e.g., within Annex I
boundaries.
3
Although if RIL makes logging more profitable, more logging may result.
Climate change mitigation projects that minimize harm from logging operations (RIL
techniques) could be assumed to result in no or little market leakage since RIL essentially
seeks to maintain a given timber output but with less damage. Differently, Brown et al. (1997)
derived a considerable leakage potential of RIL techniques based on data of a carbon-offset
project in Sabah, Malaysia. This project caused leakage because on one-third of the project
area, timber production initially decreased by approximately 50m3 per ha relative to
conventional logging. The total timber shortfall was set at 22,050 m3. They calculated the
maximum leakage potential by estimating the additional area that must be logged to make up
for this deficit. Assuming that RIL techniques would be used to compensate this shortfall,
they arrived at an offsite carbon emission of 23,112 t C, equal to 60% of the annual gross
carbon benefits of the project (38,700 tC). This maximum figure is useful in demonstrating a
simple means of quantifying potential leakage. The true leakage effect of this project will
most likely be less since over time RIL should increase timber output compared to
conventional logging by sustaining less damage to young trees.
3.2 Reforestation and afforestation
Tree-planting projects run the gamut from diversified, community-based agroforestry systems
that meet local needs, to restoration of degraded grazing land, to large-scale, commercial
monocultures that supply global markets. Without adequate empirical data, it is difficult to
say whether certain types of sequestration project may be more or less leakage prone. In
general, small-scale reforestation/afforestation projects focused on environmental restoration
and local community needs should have more positive leakage profiles than large-scale
commercial plantations, mainly because they aim to minimize the drivers and scale of
potential leakage.
If carried out on currently productive land, reforestation or afforestation may cause activityshifting leakage as displaced people migrate to other forest areas. Displacement of local
communities from lands has been a frequently cited concern, especially in the case of largescale commercial reforestation projects (World Rainforest Movement 1999; Sawyer 1993). In
Honduras, for example, thousands of small farmers and ranchers were reportedly displaced
from the north coast valleys to make way for the establishment of oil palm cooperatives
(CFAN 2001). While oil palm plantations would most probably not be considered forestry,
under the Kyoto Protocol, this case demonstrates the enormous displacement potential of
large-scale mono-component land-use projects.
As a rule, activity-shifting leakage will depend on whether reforestation engages or displaces
landowners. If reforestation is an alternative or complementary land use for existing
landowners, and economic benefits are comparable to non-forest alternatives, then the risk of
negative activity-shifting leakage will likely be low.
Projects that yield valuable
environmental services may lead to additional carbon and other benefits, outside the project
boundary (e.g., positive leakage). One example would be an agro-forestry project that
improves water quality and enhances agricultural production, leading to additional
communities practicing agro-forestry.
Market leakage from reforestation/afforestation projects can potentially be either positive or
negative. In certain circumstances, however, reforestation/afforestation projects will be
virtually free of market leakage risk. This would be the case for projects that serve
exclusively to sequester and store carbon for the long-term and/or produce other
environmental services, such as restoration projects for watershed protection or biodiversity
conservation, on previously degraded lands. Second, if a reforestation/afforestation project
generates products that are substitutes for others that come exclusively from natural forest
sources (e.g. firewood for local use), this should tend to produce positive market-based
leakage by creating a new supply of locally available resources.
Large-scale timber plantations may be particularly prone to market leakage. They could
depress timber prices by increasing supply, which in turn could reduce the incentive to
establish new timber plantations elsewhere. This could lead to either positive leakage (if the
“foregone” plantations would have replaced carbon-rich native forests), or negative leakage
(if the “foregone” plantations would have taken place on degraded lands with a lower carbon
content). Equally plausible, timber plantations might reduce timber prices and lessen
harvesting on other natural forests. Whether market forces will lead to a decline in forest
harvesting (positive leakage) or the abandonment of plantations (negative leakage) will
depend on local market conditions and circumstances. Reforestation/afforestation projects
also may lead to negative market leakage if they reduce agricultural or livestock output or
increase agricultural land rents.
Although still relatively limited, virtually all research on the amount of leakage has been done
on market leakage from commercial plantations projects. The most advanced study in this
field, done by Sohngen and Sedjo (2000), is based on modeling a high additional input of 50
million hectares (ha) of carbon plantations established over a 30-year period. This is roughly
equivalent to a doubling in the current rate of plantation establishment.4 The general result of
this model is that potential leakage from this program could be considerable. This study
suggests that 50 million ha of new carbon plantations would decrease sequestration outside
the new forests by around 50%. This effect is largely due to accompanying adverse changes
in age classes of trees and increased management intensity rather than a diminished incentive
to establish new plantations.5
An even more drastic result was derived by Alig et al. (1997). They found a leakage rate for
carbon sequestration projects exceeding 100% following a 4.9 million hectares afforestation
program in the U.S. Their result is driven by an increase of agricultural land rents that results
in a more than one-to-one conversion rate of forestry to agriculture. Afforested lands in this
study were not required to remain in forestry (i.e. permanent set-asides) but also exposed to
economic pressures from increased agricultural land and output prices.
The Cintrafor Global Trade Model (CGTM) (Perez-Garcia 1994) resembles Alig et al.’s
work in some respects, but analyzes the impacts of a U.S. tree planting programs on the forest
sector using an economic model of world forest products markets. This study examines
impacts of a large tree-planting program in the southern U.S. on the forested land base in the
U.S.'s north and west regions as well as Canada (regional effects), and on U.S. forest products
trade with countries around the Pacific Rim and with European markets (global effects). The
local effects in this study are found to be significant but there is almost no perceivable global
impact from this tree-planting program due to existing trade links.
Kadekodi/Ravindranath (1997) show in a separate study that a nationwide teak tree-planting
program in India, while tripling the output value of forest products and reversing the forest
product trade in favour of India, would also reduce global teak prices. Leakage results from
4
The current worldwide rate of newly established of plantations is 1.6 Mio per year according to FAO (FAO
2000). If simply extrapolated it is equal to 48 Mio. ha over a 30 years period .
5
The 50 million hectars impulse of this study would reduce land areas in commercial (i.e. “non-carbon”)
plantations only by 0.2 - 7.8 million hectares (less than 15%).
this program since the global price decline would cause existing teak planters to face a smaller
return on their investments.
Table 3: Magnitude of LULUCF Market Leakage
4
CDM
Activity
Plantations
Study
Plantations
Alig/Adams/McCarl/
Callaway/Winnett 1997
Top-down 50 years
100%
Plantations
Perez-Garcia 1994
Top-down 50 years
Plantations
Top-down 50 years
Averted
deforestation
Kadekodi/Ravindranath
1997
Sohngen/Mendelsohn/
Sedjo 1999
Regionally
high, but no
global
impact
Non
negligible
Little or no
perceivable
impact
Reduced Impact
Logging
Brown/Cabarele/
Livernash, 1997
BottomUp
Sohngen/ Sedjo 2000
Study
Study
Leakage
Method
Period
Potential
Top-down 100 years6 50%
Top-down 125 years
1 year
OPTIONS FOR RESPONDING TO LEAKAGE:
POLICIES
60%
Driving factors
Increased timber
supply, adverse
changes in age
classes of trees
and forest
management
intensity
Rising
agricultural
land rents
Increased timber
supply
Increased
timber supply
Intensification
of existing forest
management &
additional
plantations in
subtropics
Maximum
leakage potential
based on a
complete
substitution of
decreased timber
production
PROJECT LEVEL RESPONSE AND
Leakage is a potentially significant risk for project activities. But as well-designed pilot
projects around the world have demonstrated, leakage is not insurmountable. We divide the
tools used to reduce and manage negative leakage in LULUCF projects into project-level
approaches and macro-level approaches.
4.1
6
Project-level options for reducing and managing leakage
Figures do not change considerably when looked at a shorter period of 50 years.
At the project level the two main categories for addressing leakage are project-specific
approaches and standardized approaches. Project-specific approaches address case-by-case
local circumstances such as fuel wood scarcity or ecosystem characteristics. Standardized
approaches are broad measures applied across classes of projects to compensate for leakage.
For example, a standardized “leakage discount” may apply to carbon offsets generated in
forest conservation projects reflecting average leakage rates for this class of projects.
Approaches at the project level taken in various projects to date are summarized in Table 4
and then discussed individually.
Table 4: Leakage Control and Monitoring in LULUCF Projects (*)
Project Specific Approaches
Policies
Site
selection
Multicomponent
project
Off site
monitoring
Discounting
Aggregate
baselines
Noel Kempff,
Bolivia
-
a
a
a
-
-
Rio Bravo,
Belize
-
a
-
a
-
-
CARE,
Guatemala
-
a
-
-
-
-
Krkonose,
Czech Rep.
a
-
-
-
-
-
Ecoland, Costa
Rica
a
-
a
-
-
-
Costa Rican
National Parks
-
a
-
-
Scolel Te,
Mexico
-
a
-
a
a
-
a
-
RIL/Sabah,
Malaysia
-
-
a
-
-
-
Projects
Leakage
Standardized
Approaches
Contracts
*) Compiled from Noel Kempff Technical Operating Protocols 1999, Brown et al. 1997 (Rio Bravo, CARE,
Krkonose), Trexler et al. 2000 (Ecoland), Chacon et al. 1999 (Costa Rican National Parks) and E-Mail
communications with Richard Tipper (Scolel Te) and Francis Putz (RIL/Sabah).
4.1.1
Project-specific approaches
4.1.1.1 Site selection
At the project level, sound leakage management begins during the design of a project.
Projects should be structured to maximize positive leakage (increased sequestration, reduced
emissions) and minimize negative leakage (decreased sequestration, increased emissions). An
obvious way to do this is for project developers to carefully seek out and consult communities
with sincere interest in enhancing management of their forestlands. Thoughtful site selection
might also entail locating a project in an area with few or no competing uses (such as
degraded lands) or limited access (such as remote forests threatened with new logging roads)
to minimize the risk of displacing people or causing negative market leakage. An example of
a careful site selection is the Krkonose reforestation project in the Czech Republic where the
project developer chose an isolated location with virtually no risk of encroachment or
displacement (Brown et al. 1997). Another example is the Ecoland forest conservation project
in Costa Rica, where virtually no forest was left standing outside the project area, making
activity-shifting leakage improbable (Trexler et al. 2000, p. 147).
4.1.1.2 Multi-component projects
Projects should create incentives for local people to maintain the project and its GHG benefits
by providing a range of socio-economic benefits (CIFOR 2000). Multi-component projects
may help realize this and avoid leakage by integrating measures to meet local needs and
provide sustainable access to resources (timber, fuel wood, cropland, etc). For example,
project developers of the Noel Kempff Mercado forest conservation project in Bolivia control
leakage with demand-management activities. These mitigation activities include agroforestry
to provide sustainable sources of wood, employment opportunities, and equipment retirement
schemes (Noel Kempff Technical Operating Protocols 1999). Similarly, the CARE Guatemala
project increased fuel wood supply and agricultural productivity by providing trees through
tree nurseries (Brown, et al. 1997). These components minimize negative activity shifting and
negative market leakage (less pressure on the original forest) while encouraging additional
positive ecological leakage. Other advantages to multi-component design include: possible
enhanced profitability, increased local employment, numerous ecological positive feedbacks
and overall higher likelihood of community support and success (Niles and Schwarze 2000).
4.1.1.3 Leakage Contracts
The Noel Kempff Mercado Climate Action Project forest conservation project in Bolivia uses
contracts that prohibit leakage. The logging concessionaires bought out by the project signed
contracts committing them to technical assistance and training in sustainable forest
management practices for their remaining concessions (supporting positive activity-shifting
leakage). To decrease activity shifting, concessionaires also agreed not to use money received
from the project to purchase new concessions and to abandon logging equipment onsite (Noel
Kempff Technical Operating Protocols 1999). To monitor possible activity shifting by
loggers, the Noel Kempff Mercado Climate Action Project is scheduled to follow the
investments and actions of timber concessionaires who were displaced by the project. Other
contractual approaches might require (and possibly compensate) farmers to plant trees or use
more efficient cooking stoves. It is likely that enforcement of leakage contracts will be easier
with a smaller number of stakeholders.
4.1.1.4 Monitoring
All types of climate mitigation projects and policies should monitor leakage to a degree
commensurate with the risk and possible scale of leakage. Since project design will not
always be able to prevent leakage altogether, careful monitoring and measurement can
provide the basis for adjusting GHG benefits accordingly.
There are several ways to monitor for potential leakage. This may involve expanding the
geographic scale of monitoring beyond project boundaries to capture regional effects through
a combination of remote sensing and sample plots in non-project areas. This method has been
explored in the Scolel Te Project in Mexico (Tipper and de Jong 1998). Although expanding
the boundaries for monitoring may capture activity-shifting and market leakage effects, it is
often difficult to isolate impacts of project leakage from other exogenous factors that affect
markets and land use (e.g. demographics, access, policies, markets for agricultural products or
a host of other factors). The definition of boundaries for monitoring and the interpretation of
results thus represent a special challenge.
The scope of monitoring may also be expanded to examine market linkages beyond the
immediate project vicinity. Project managers can try to track effects of a project on the
supply, price and sources for timber or agricultural products. However, markets are hard to
model and are notoriously difficult to predict, even under the best of circumstances.
Analyzing markets for timber, crops or livestock in many developing countries is complicated
by the predominance of the informal sector and the lack of accurate and current data.
Another approach, proposed by Brown et al. (1997), is to track selected indicators rather than
monitoring leakage directly. The use of indicators can alert monitoring agents as to the risk of
leakage. This approach emphasizes key indicators of demand driving baseline land-use
change. If demand for timber, fuelwood or farmland is left unmet due to a project, this would
signal a risk of leakage. This in turn could trigger a burden of proof for project management
of no-leakage (or any other type of so-called ‘seller liability’) similar to the traffic-lightapproach for non-compliance (Wiser and Goldberg 1999). The indicator approach has the
advantage of being relatively straightforward and transparent. Key indicators may be defined
beforehand and can be objectively evaluated over time. Such an approach is likely to be more
qualitative than quantitative. It remains a challenge to precisely measure how a decreased
supply of farmland or forest products translate into carbon leakage.
Though laborious, tracking activities of resource users affected by project activities may be
an effective means for capturing activity-shifting leakage. Activity shifting will often be
easier monitored at the local level (although, for example, some logging companies could
migrate to other countries) and it may be more tractable at the project scale. A variety of
social and economic assessment methods, including surveys, follow-up interviews, and
review of land transactions, can be used to estimate activity-shifting leakage. As mentioned
prior, the Noel Kempff Mercado Climate Action Project follows investments and actions of
timber concessionaires displaced by the project.
4.1.2 Standardized approaches
4.1.2.1 Discounting
If leakage is effectively monitored and calculated, then greenhouse gas benefits from a project
may be adjusted accordingly. If projects monitor leakage in detail, any discount can be based
on the specific outcomes of project implementation and monitoring. As previously illustrated,
however, monitoring and calculating all possible leakage effects will be practically
impossible. In order to address leakage from a practical standpoint and to simplify leakage
accounting, standard coefficients could be used to adjust the GHG benefit estimates of
projects. Some observers (Lee et al. 1997) have suggested applying a standardized discount to
LULUCF projects. These adjustment coefficients will vary for different project types and
different national or regional circumstances (IPCC 2000, p. 314). Case studies (IPCC 2000,
Sec. 5.3.3.2., p. 264) indicate that certain broad landscape characteristics could be used to
establish these varying leakage discounts. For example, a plantation project on severely
degraded lands with few alternative uses and no occupants will almost certainly be less prone
to leakage than a plantation carried out on productive lands with many users and/or
occupants. Establishing appropriate adjustment coefficient(s) to cover diverse leakage risks
merits further research, but ultimately will require policy decisions that balance the needs for
accuracy and workability.
4.1.2.2 Project Eligibility Criteria
A drastic form of “discounting” would be to rule out projects that are perceived as particularly
leakage prone. For example, if forest conservation projects in developing countries are set
non-eligible for CDM projects (as under the current Art. 12 ruling of COP 6-bis) potential
carbon credits from these activities are effectively discounted at a rate of 100%.
Given that leakage will mostly be project-specific and can be addressed through careful
project-design, removing whole suites of objects would risk eliminating projects that yield
positive GHG benefits and net local benefits. A favorable alternative approach would be to
grant projects with appealing leakage profiles a streamlined approval and monitoring process
similar to the preferential treatment of small-scale renewables and energy efficiency CDM
projects decided at COP6-bis (UNFCCC 2001, 9).
4.1.2.3 Aggregate baselines
Another way of expanding a project’s accounting boundary is to use aggregate baselines. This
technique would entail developing national, regional or sectoral baselines on land-use change
and management. Experience with standardized baselines, however, has shown that
calculating national or sector-type baselines brings in tertiary and other indirect effects that
can overwhelm any attempt at project-caused calculations (Trexler 1999, p.46). This problem
is further aggravated by the fact that land-use change data for most developing countries is
very incomplete. For example, deforestation rate estimations can typically vary greatly for
the same country according to different sources and methods (Mathews 2001). Aggregate
baselines are also rejected by many developing countries on the grounds that such national
procedures could be a “back-door” way to coerce developing countries into a regime of
quantified emission reduction targets.
4.2
Macro-Level Options for Reducing and Managing Leakage
At the macro level, we identify two major proposals that can minimize LULUCF leakage.
First, a ceiling or cap can be placed on the volume of forestry projects for climate change
mitigation. This approach serves to minimize the perceived scale of leakage risks. The second
option is a balanced portfolio approach. It aims to offset, or balance, market leakage effects
by combining reforestation/afforestation and avoided deforestation projects at the policy
level.
4.2.1
The 1%-ceiling on CDM sinks
The resumed session of the Sixth Conference of the Parties, Part 2 (COP-6 bis) limited the
number of credits developed countries may gain via CDM forestry projects. During the first
commitment period (2008-2012), developed countries may only use CDM forestry projects to
meet 1% of their respective 1990 emissions per year, for each of the five years in the first
commitment period (UNFCCC 2001). Furthermore, at COP-6 bis, a decision was taken to
restrict the type of forest projects to those that involve reforestation and afforestation
(“sinks”) and limited the size of individual projects. Although these restrictions were not set
primarily to address the risk of leakage, limiting the scope and number of projects may
effectively keep in check the impact of projects on local and/or global timber markets
respectively.
The COP-6 bis ceiling will probably not have a great impact, as it appears to have been set
above the threshold likely to be reached by CDM reforestation and afforestation projects.
These results are discussed in a separate forthcoming paper (Niles and Vrojlik 2002)
4.2.2
A balanced LULUCF portfolio
Plantations may have opposite effects on local or global timber markets compared to forest
conservation projects (see figure 2). Reduced logging of natural forests decreases the supply
of forest products (price increase) and accelerated planting of trees increases the supply (price
decrease). Some authors (e.g. Chomitz 2000, p.7) have argued that these counter-acting forces
could lead to a leakage-neutralizing situation, where market effects of halting deforestation
offset opposite market effects of plantations and/or reforestation. While there is no empirical
evidence to support this notion, basic economic rational suggests this possibility may have
some merit. The possibility of such “counteracting” forces makes an argument for trying to
“balance” the ratio of these projects. This would imply that the recent decision by policy
makers to exclude forest conservation from forest projects in developing countries could
exacerbate leakage. Thus, the CDM executive board will not have this policy tool at its
disposal, at least in the first commitment period.
Price
Figure 2: Market effects of LULUCF projects
Demand
Supply with reduced logging
Supply
Supply with increased aforestation
1
2
1= Reduced logging of virgin forests
2= Increased afforestation of abandoned land
Quantity of Forest Products
Such a balanced-portfolio approach, while useful to some degree, assumes a similarity in
market effects of tree planting and tree protection that may not exist. First, industrial
plantations operate on different time-scales compared to forest conservation projects. It can
take 10 years or more for plantations to increase the wood supply whereas projects that stop
logging reduce timber supplies in the short term and decades into the future. Interestingly, a
relative symmetry could be achieved if avoided deforestation was allowed in the 2nd
commitment period when new timber plantations will start being harvested. Second, harvest
intensities vary greatly among (high yielding) industrial plantations and (low yielding)
logging of tropical forests. There is also substantial variability within similar project types
across regions and ecosystems (IPCC 2000, chapter 5). A balanced portfolio approach would
need to take these differences into account, something not practical on a simple per-project
basis. Finally, logging is only one factor behind tropical deforestation so carbon conserved
through forest protection may only be loosely related to its effects on timber markets. The
“balancing” approach therefore seems more promising at the project level, as discussed in
section 4.1.1.2.
4.3 A Decision-Tree Approach for Managing Leakage
No approach to addressing leakage will be perfect. Methods will have to balance the need for
accuracy with the desire for transparency and simplicity. Developing tailor-made methods for
measuring and accounting leakage on a project-by-project basis may result in greater seeming
accuracy. However, a profusion of approaches will undoubtedly be more opaque and subject
to gaming—not to mention more costly and complex. Standardized approaches such as
discounting may be less accurate, but their criteria can be conservatively defined and will be
far less subject to gaming.
Several guidelines might serve for dealing with leakage in a transparent and uniform way
across projects. Aukland et al (2002) lay out a conceptual “decision-tree” framework for
describing in qualitative terms the potential leakage dynamics of a project. This framework
could be combined with project-specific and/or standardized factors to make quantitative
adjustments or discount project GHG benefits. A sample decision tree employing these
guidelines for an avoided deforestation project, modified from Aukland et al (2002), is
included in Appendix 1.
In conservation projects, activity-shifting and market leakage for conservation projects have
partially countervailing effects. If activities simply shift place, market leakage will be reduced
since the net change in the output of goods will decline. Total leakage for avoided
deforestation projects is therefore the sum of activity-shifting leakage and market leakage
adjusted for the effect of activity-shifting leakage.
TOTAL LEAKAGE = (LAS+ ((100- LAS) x LM) * GPB
Where:
LAS = Activity shifting leakage (%)
LAM = Market leakage (%)
GPB = Gross Project Benefits (tC or tCO2)
For reforestation and afforestation projects, activity shifting does not necessarily mitigate
market leakage as for conservation projects. A commercial reforestation project could
simultaneously displace baseline agricultural activities into new forest areas (activity-shifting
leakage) and depress investments in new forestry plantations by increasing supply (market
leakage). In this case, total leakage is therefore the sum of activity-shifting leakage and
market leakage.
TOTAL LEAKAGE = (LAS+ LM) * GPB
Reforestation projects can also lead to less harvesting in natural areas (more timber, more fuel
wood), whereby activity shifting would be negative and market leakage would be positive.
5
CONCLUSIONS AND RECOMMENDATIONS
Unintended consequences arise in virtually every type of activity. Climate change mitigation
and sustainable development are no different. Discrete projects or policies may cause impacts
that extend over a broad scale, through complex and dynamic causes and effects.
Devising solutions and rules for leakage is hampered by a dearth of substantial quantitative
studies. There are no long-term, peer-reviewed evaluations of climate change mitigation
leakage for actual projects. Until more studies are conducted, leakage is a legitimate concern
in terms of its magnitude, although results from studies are highly variable. They also may not
turn out to resemble real leakage rates that eventually are measured. There is no evidence that
forestry as a sector is more prone to leakage than are other sectors, such as energy or
transportation (although there are some indications for this belief). Nor is there concrete
evidence that any one type of forestry project is more or less susceptible to leakage than
others (although there are reasons to suspect that plantations may be particulary prone to
market leakage).
Leakage will probably be most determined by project-specific activities, not broad categorical
groupings. Together, these findings suggest that all climate mitigation projects – plantations,
transportation, forest conservation and energy - should monitor for and address leakage at
the project level. There is no apparent and compelling reason to shun any one type of climate
change mitigation based solely on leakage.
Notwithstanding the lack of data, modeling scenarios, case studies, and general observations
can provide insight into preventing negative leakage and fostering positive leakage. Projects
should try to maintain the same long-term output of products (or substitutes) of baseline
activities to account for market leakage. Multi-faceted projects may have advantages in terms
of minimizing negative leakage and maximizing positive leakage. Integrated projects that use
reforestation, plantations and forest conservation programs, seem well suited to create
sustained, diverse economies that will not cause severe market disruptions or unwanted
migration or displacement. Project diversity should also extend to non-forestry dimensions,
such as clean energy, reliable transportation and other aspects. Other instruments, such as
wise site selection and leakage contracts, may also help prevent negative leakage and bad
projects.
At regional, national or global scales, leakage can also be addressed with complimentary
project types. There is some evidence, for example, that reforestation programs and forest
protection programs could have balancing influences on the global timber market. However,
important asymmetries between growing new forests and not cutting down old ones must be
addressed (timing of impacts, per area timber effects) for the value of multiple project
activities to scale up.
Ensuring the integrity of climate change policies will also entail measuring leakage when it
does occur. Monitoring will allow appropriate adjustments to be made to the amount of
carbon offsets claimed by a project. Yet, measuring leakage with a high degree of precision
will often be costly and inconclusive. The boundaries defined, the mechanisms posited and
the assumptions employed will all be subject to contradicting opinions. It is not so much a
problem of not having ways to map and measure leakage—a host of economists, sociologists,
geographers and foresters can develop plausible approaches for each project. The problem lies
precisely in this variety—and the resulting variety of possible outcomes.
Several approaches are available for managing leakage that does occur in an affordable and
sufficiently accurate manner. First, monitoring beyond the project boundaries for selected
indicators of leakage is one practical solution. Second, discount factors may be applied in the
short term. They should be conservative (i.e., high) as the CDM market becomes operational,
and can be refined as the market matures and further information becomes available. Wellreasoned discount coefficients will be different for different regions, project types and/or
markets. For activity-shifting, this discount might be based on the likelihood of engaging
local people. For markets, elasticities of demand and supply will be useful parameters for
estimating leakage (cp. Chomitz 2000, p.8/9). Projects that have an appealing leakage profile
– that minimize negative undesirable unintended consequences while promoting positive ones
– could also be granted a preferential treatment in the proces of approval and monitoring to
reward efforts in project-design.
It is also worth recalling that many projects will potentially have positive spillover effects.
Positive activity shifting leakage (e.g. broader adoption of project technologies), market
leakage (e.g. reduced pressure on natural forests as timber sources) and ecological leakage
(ecosystem stabilization) may end up on par with negative leakage. These positive and
unaccounted GHG benefits may also “buffer” against negative leakage that is not captured by
monitoring.
ACKNOWLEDGMENTS
The authors would like to recognize the generous contributions of information and comments
on earlier drafts of this paper by Brent Sohngen, Duane Lakich-Muelller, John Perez-Garcia,
Richard Tipper, Marilyn A. Buford, Bryce Stokes, Kenneth Andrasko, Toral Patel-Weynand,
Paige Brown, Bill Stanley and Eric Firstenberg. The opinions expressed, and any possible
errors, are of course solely our responsibility.
1
REFERENCES
Alig, R., Adams, D., McCarl, B., Callaway, J.M., and S. Winnett. 1997. Assessing Effects
of Mitigation Strategies for Global Climate Change with an Intertemporal Model of the U.S.
Forest and Agricultural Sectors. Environmental and Resource Economics 9: 269-274.
Aukland, L., P. Moura Costa and S. Brown. 2002. A conceptual framework for addressing
leakage in avoided deforestation projects.
Brown, P., Cabarele, B. and R. Livernash. 1997. Carbon Counts: Estimating Climate
Mitigation in Forest Projects. World Resource Institute, Washington D.C.
Chacon, C., Castro, R. and S. Mack. 1998. Pilot Phase Joint Implementation Projects in
Costa Rica. A Case Study. Center for International Environmental Law, Washington D.C.
http://www.ciel.org/Publications/CDMCaseStudy.pdf. Accessed Nov 2001.
Chomitz, K.M. 2000. Evaluating Carbon Offsets from Forestry and Energy Projects: How
Do They Compare?
World Bank Paper 2357, World Bank, Washington D.C.
http://econ.worldbank.org/docs/1111.pdf.
Center for International Forestry Research (CIFOR). 2000. Capturing the Value of
Forest Carbon for Local Livelihoods. Policy Brief from a workshop organized by CIFOR
and the University of Maryland at Como/Italy, February 2000 (distributed at COP5).
CIDA Forestry Advisers Network (CFAN). 2001. Forestry Issues. Deforestation: Tropical
Forests in Decline. http://www.rcfa-cfan.org/ english/issues.12.html.
Food and Agriculture Organization of the United Nations (FAO). 1998. Global Fibre
Supply Model. FAO, Rome.
Goldberg, D. 1999. Carbon Conservation. Climate Change, Forests and the Clean
Development Mechanism. Center for International Environmental LAW, Washington D.C.
http://www.ciel.org/Publications/CleanDevelopmentMechanismText.pdf.
Intergovernmental Panel on Climate Change (IPCC). 2000. Land Use, Land Use Change,
and Forestry, edited by Watson, R.T., Noble, I.R., Bolin, B., Ravindranath, N.H., Verado,
D.J., and D.J. Dokken, Chapter 5, Project Based Activities, edited by Brown, S., Masera, O.
and J. Sathaye
http://www.grida.no/climate/ipcc/land_use/index.htm.
Kadekodi, G.P. and N.H. Ravindranath. 1997. Macro-Economic Analysis of Forestry
Options on Carbon Sequestration in India. Ecological Economics 23: 201-223.
Laurance, W. F., Laurance, S. G., Ferreira, L. V., Rankin-De Merona, J. M., Gascon,
C., Lovejoy, T. E. 1997. Biomass Collapse in Amazonian Forest Fragments. Science 278:
1117-1118.
Lawton, R.O., Nair, E.S., Pielke, R.A. and R.M. Welch. 2001. Climatic Impact of Tropical
Lowland Deforestation on Nearby Montane Cloud Forests. Science 294: 584-587.
Lee, R. 1997. Understanding Concerns About Joint Implementation. Joint Institute for
2
Energy and Environment/Oak Ridge National Laboratories, Oak Ridge/TN.
Mathews, E. 2001. Understanding the FRA 2000. Forest Briefing No. 1, World Resources
Institute, Washington D.C. http://www.wri.org/wri/pdf/fra2000.pdf.
McGuffie, K., Henderson-Sellers, A., Zhang, H., Turbridge, T.B., and A.J. Pitman, 1995.
Global Sensitivity to Tropical Deforestation. Global Planet Change 10: 97-128.
Niesten, E., Hardner, J., and R.E. Gullison. 2001. Inter-Annex Carbon Leakage Related to
Forestry. Paper prepared for the Union of Concerned Scientists, February 2001.
Niles, J.O. and R. Schwarze. 2000. Long-Term Forest Sector Emission Reductions and
Article 12, in: Robertson, K.A. and B. Schlamadinger, B. (eds.), Proceedings of the IEA
Bioenergy Task 25 Workshop: Bioenergy for Mitigation of CO2 Emissions. Gatlinburg/TN,
p. 145-155.
Niles, J.O. and C. Vrojlik. 2002 The Scale of Land Use, Land-Use Change and Forestry
(LULUCF) in Developing Countries for Climate Mitigation. The Nature Conservancy,
Washington D.C..
Noel Kempff Climate Action Project. 1999. Technical Operating Protocols.
Perez-Garcia, J. 1994. An Analysis of Proposed Domestic Climate Warming Mitigation
Program Impacts on International Forest Products Markets. CINTRAFOR Working paper
50. http://www.cintrafor.org.
Rowe, R., Sharma, N. P. and J. Browder. 1992. Deforestation: Problems, Causes and
Concerns, , Pp. 33-45vin: Sharma, N.P (ed.), Managing the World’s Forests: Looking for
Balance Between Conservation and Development. Kendall/Hunt Publishing Co., Iowa..
Sawyer, J. 1993. Plantations in the Tropics: Environmental Concerns. IUCN/UNEP/WWF,
Gland/Switzerland.
Schwarze, R. and J. O. Niles. 2000. The Long-Term Requirement for Clean Development
Mechanism Forestry and Economic Liability. Journal of Environment and Development 9:
384 - 404.
Schimel, D, Alves, D., Enting, I., Heimann, M., Joos, F., Raynaud, D. and T. Wigley.
1996. Radiative Forcing of Climate Change, Pp.65-132 in Houghton, J.T., L.G. Meira
3
APPENDIX 1: ESTIMATING LEAKAGE FROM AVOIDED
DEFORESTATION PROJECTS
(BASED ON DECISION-TREE APPROACH OF AUKLAND ET AL 2002)
STEP 1: Disaggregate
sources of GHG benefits
Estimate relative contribution of
baseline activities to estimated
GHG benefits
(Develop following steps for
each source of GHG benefits)
•
STEP 2: Calculate
Activity-Shifting Leakage
•
•
Forests minimal portion of
regional land use, or
Barriers effectively limit
possibilities for relocating
activities replaced by project,
or
Activities and stakeholders
continue within project
YES
NO
Activity shifting leakage
(LAS) project = 0
100%
engagement
Project includes components that
engage stakeholders in
alternative activities
<100% engagement
Activity-Shifting Leakage(LAS) =
proportion of baseline
activity/stakeholders not engaged in
alternative activities * proportion of
region under forest cover
STEP 3: Calculate
Market Leakage
4
ESTIMATING LEAKAGE FROM AVOIDED DEFORESTTION
PROJECTS
2
Project affects long-term supply
of goods to markets
YES
Project includes measures to
reduce demand or increase
supply of goods or substitutes
STEP 3: Calculate Market
Leakage
NO
Consider market leakage(LM) = D
Market leakage (LM) from project = 0
YES
YES
Measures effective, no net effect
on markets
NO
Market leakage(LM) = %
reduction in baseline supply x D
STEP 4: Calculate
Total Leakage
TOTAL LEAKAGE = (LAS+ ((100- LAS) x LM) * GPB
Where,
LAs = Activity shifting leakage (%)
LAM= Market leakage (%)
D = Standard adjustment coefficient for market leakage by project type and/or region(%)
GPB = Goss Project Benefits (tC or tCO2)
5